Antidepressants and suicide: Authors' reply

doi:10.1136/bmj.329.7463.461-b

BMJ 2004;329:461-462 (21 August), doi:10.1136/bmj.329.7463.461-b

Letter

Antidepressants and suicide

Authors' reply

EDITOR—We agree with Aldred and Healy's suggestion thatpart of the recent rise in antidepressant prescribing may bedue to there being a growing number of long term users of thesedrugs—we acknowledged this as a limitation of our model.If any adverse effects of antidepressants on suicide risk occurmainly in the first few weeks of treatment then our model willoverestimate these.

Mitchell correctly identifies that one of the major assumptionswe made in our model was that the risk estimates of non-fatalsuicidal behaviour derived from paediatric trials could be appliedto fatal suicidal behaviour in adults. We acknowledged thisimportant limitation of our modelled "worst case scenario" inthe paper. Mitchell points out the drug specific risk estimateswe reported in the table are risk ratios rather than odds ratios.The odds ratios for the drug specific estimates are very similarto the relative risks. Mitchell may have misunderstood one aspectof our model by suggesting it estimates the excess of non-fatalsuicidal acts. We derived our estimates from prescribing data,the suicide rate among patients receiving antidepressants inprimary care,¹ and we assumed (see above) that the relativerisk of non-fatal suicidal behaviour in paediatric trials ofselective serotonin reuptake inhibitors (SSRIs) is similar forsuicide and all ages. If we had wished to estimate effects onnon-fatal self-harm we would have used the rate among peoplereceiving antidepressants rather than the suicide rate; as ratesof non-fatal self harm are over 20 times higher than those forsuicide this would result in a higher estimate.

David Gunnell, professor of epidemiology

D.J.Gunnell{at}bristol.ac.uk Department of Social Medicine, University of Bristol, Bristol BS8 2PR

Deborah Ashby, professor of medical statistics

Wolfson Institute of Preventive Medicine, Barts and London, Queen Mary School of Medicine and Dentistry, University of London, London EC1M 6BQ

Competing interests: DG and DA are members of the Medical andHealthcare Products Regulatory Agency's expert working groupon the safety of SSRIs, and DA is a member of the Committeeon Safety of Medicines. Both act as independent advisers, receivingtravel expenses and a small fee for attending meetings and readingmaterials in preparation for the meeting. DA has spoken on themethods of adverse drugs reactions in HIV at a scientific meetingattended by several pharmaceutical companies, and sponsoredby GlaxoSmithKline. An honorarium was paid to her department.